LibriSpeech-2Mix

LibriSpeech2Mix is a dataset used to evaluate multi-speaker automatic speech recognition (ASR) systems, containing simulated English speech mixtures for assessing performance in overlapping conversations.

Papers using LibriSpeech-2Mix (3)

Conformer-based Target-speaker Automatic Speech Recognition For Single-channel Audio2023 · 17 citesCALM: Joint Contextual Acoustic-Linguistic Modeling for Personalization of Multi-Speaker ASR2026Conformer-based Target-Speaker Automatic Speech Recognition for Single-Channel Audio2023